| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Tumor Biology |
Department of Biological Regulation, Weizmann Institute of Science, Rehovot, 76100 Israel [H. D., T. I., M. N.]; Department of Radiology, Johns Hopkins University School of Medicine, Baltimore, Maryland 21205-2195 [Z. M. B.]; and Division of Cancer and Angiogenesis, Department of Pathology, The Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, Massachusetts 02215 [L. E. B.].
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
VEGF-C/VEGFR-3 signaling may increase the surface area of the lymphatic vasculature and consequently increase the probability of lymphatic invasion by tumor cells. Tumor interstitial hypertension, resulting from increased permeability of the neovasculature, was also proposed as a major determinant of metastatic spread (12, 13, 14) . The most potent signaling for vascular permeability is provided by VEGF, also known as vascular permeability factor (7 , 15) . Increased interstitial pressure is expected to activate lymphatic drain by pulling the anchoring filament attached to the lymphatic endothelium and opening pores in the vessels walls (16) . Therefore, fluid leaking out from a solid tumor may convey with it cells and molecules, and flow through the tissue, down the pressure gradient, and into the draining lymphatics.
The goal of this study was to investigate the effects of tumor vascular permeability on interstitial convection and lymphatic uptake in the peritumor tissue. Using MRI and confocal microscopy we showed previously that acute exposure of the skin vasculature to a bolus of recombinant VEGF165 resulted in vasodilation and transient increase in vascular permeability (17) . In addition, some of the extravasated macromolecular contrast material was drained into lymphatics (17) . In the present study, we applied novel MRI tools for monitoring lymphatic drain from C6 glioma tumors that were engineered to overexpress VEGF165 under switchable tet-off regulation (C6-pTET-VEGF; Refs. 18 , 19 ). Albumin triple-labeled with biotin, GdDTPA, and fluorescein was used as a multimodality contrast material. After i.v. administration, blood-plasma volume and permeability maps were derived by MRI, and confirmed by fluorescence microscopy and histology. Avidin chase (20) was then applied for inducing rapid clearance of the intravascular contrast material, providing experimental separation between vascular leak and lymphatic drain. Vascular permeability induced by VEGF is shown here to produce a driving force for peritumor convection and lymphatic drain of macromolecules that extravasated from the tumor vasculature. These findings implicate VEGF in generation of directional convective flow that could facilitate lymphatic-mediated metastasis.
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Contrast Materials.
BSA-based macromolecular contrast material, biotin6-BSA-GdDTPA26, (Mr 
82,000; proton relaxivity, 192.9 mM-1s-1 at 4.7 T) was prepared as reported previously (17)
. An i.v. dose of 10.4 mg in 200 µl/mouse was used.
Concentration of biotin6-BSA-GdDTPA26 in the blood was evaluated in blood samples, from mice with VEGF overexpressing C6-pTET-VEGF tumors, using 4-hydroxyazobenzene-2 carboxylic acid/avidin reagent (Sigma Chemical Co., St. Louis, MO; n = 6) and from MRI of the tail veins (n = 4). Blood level was constant during the first 10 min after i.v. administration and dropped to 70% of the initial level by 60 min (Fig. 1)
. Similar pharmacokinetics were reported previously for albumin-GdDTPA (21
, 22)
.
|
MRI Experiments.
MRI measurements were conducted on a 4.7 T horizontal Bruker Biospec spectrometer using a whole body excitation coil and an actively decoupled 1.5-cm surface detection coil. Single slice spin echo images of the tumor-bearing limb were acquired as reported (Ref. 17
; TR 1,000, 500, 200, and 100 ms, TE 10.6 ms, 2 averages, spectral width 50,000 Hz, FOV 20 mm, slice thickness 1 mm, matrix 128 x 128, in plane resolution 156 µm, acquisition time 52 s for TR of 200 ms). Two-dimensional T1 weighted spin echo images were interleaved by three-dimensional T1 weighted GEs; TR 10 ms, TE 3.6 ms, flip angle 30°, 2 averages, spectral width 50,000 Hz, FOV 20 mm, matrix 128 x 128 x 64, resolution 156 µm x 156 µm x 312 µm, acquisition time 163 s). Intravascular concentration was determined from multislice spin-echo images of the tail (TR 1,000, 500, 200, and 100 ms, TE 10.6 ms, 2 averages, spectral width 50,000 Hz, FOV 10 mm, slice thickness 4 mm, matrix 128 x 128, in plane resolution 78 µm, acquisition time 10 min for each series of TRs).
Analysis of the Dynamic MR Data.
Pixel-by-pixel analysis was used to generate concentration maps of biotin6-BSA-GdDTPA26 from spin-echo data sets as reported (17)
. The concentration maps were used for derivation of three parameter maps: (a) the fPV, the ratio between the extrapolated concentration of biotin6-BSA-GdDTPA26 at time of administration and the concentration in blood-plasma (84 µM); (b) the APS, the initial rate of contrast accumulation (linear fit of the first 10 min); and (c) macromolecular convection: Time2Max maps, the time at which the rate of accumulation of biotin6-BSA-GdDTPA26 was maximal (postcontrast time was divided into 11 semioverlapping intervals of 10 min, for each interval the linear slope was derived, and slopes with r2 < 0.7 were discarded).
Changes in fPV and APS were calculated from mean concentration values at selected regions of interest including the region enhancing after the first 10 min or the tumor periphery for nonenhancing tumors. Mean values ± SD are reported.
Interstitial flow velocity was calculated from concentration maps as the average radius of enhancement ring outlined by the progression of the front of the contrast material between 10 and 60 min after contrast.
Confocal Microscopy.
Combined contrast agent fluorescein-labeled biotin6-BSA-GdDTPA26 was injected i.v. (10.4 mg/mouse; see "Contrast Materials") in tumor-bearing mice, and was allowed to circulate and extravasate for 60 min. Rhodamine-labeled BSA was injected just before tissue retrieval (3 mg/mouse). Tissue samples (skin and s.c. tumors) were fixed and scanned by confocal microscopy (Axiovert 100; Zeiss, Goettingen, Germany) as reported (17)
.
Histology.
Samples were fixed in Carnoy’s solution (6:3:1 ethanol/chloroform/acetic acid) and embedded in paraffin. Four-µm sections were stained for horseradish peroxidase-conjugated Bandeiraea simplicifolia BS-1 Isolectin and visualized with 3-aimino-9-ethyl carbasole to detect endothelial cells (19)
. Smooth muscle cells and contrast agent were stained with alkaline phosphatase-conjugated to anti-
-SMA antibody and avidin, respectively, and visualized with Fast Red by fluorescence microscopy. Alternatively, avidin-FITC was used along with -SMA staining for double fluorescence labeling of the biotinylated contrast material and vascular smooth muscle cells, respectively. Mayer’s Hematoxylin solution was used for counter staining (all of the reagents are from Sigma).
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
; Fig. 3, A–C
), whereas contrast extravasation and signal enhancement were significantly reduced in the same tumors after VEGF withdrawal (Fig. 3, D–F)
. Concentration maps for the contrast material were derived as reported previously from the precontrast R1 maps and postcontrast T1 weighted images (17)
. Linear regression was used for pixel-by-pixel fitting of changes in the biotin6-BSA-GdDTPA26 concentration maps during the first 10 min after contrast. The intercept and slope of the linear fit were used for derivation of the fPV and the APS, respectively (Fig. 3, G, H, J, and K
; Fig. 4
). VEGF overexpressing C6-pTET-VEGF tumors were highly angiogenic as manifested by the high fPV (0.0109 ± 0.0034; n = 11) and hyperpermeability (APS = 0.136 ± 0.075 µM/min; n = 11). Switching off overexpression of VEGF significantly reduced fPV (0.0062 ± 0.0045; Fig. 4A
; two-tailed paired t test, P = 0.018) and biotin6-BSA-GdDTPA26 leak (APS = 0.0333 ± 0.0211 µM/min; Fig. 4B
; two-tailed paired t test, P = 0.002).
|
|
|
. These kinetics are consistent with progressive outward interstitial flow of the labeled albumin. To visualize this nonlinearity in contrast accumulation, Time2Max maps were generated, displaying for each pixel the time after contrast administration at which the rate of contrast accumulation was maximal (Fig. 3, I and L)
. These maps resolved the time course and the spatial variance of the outward flow, or convection, of the extravasated fluid and contrast material. Hyperpermeability resulting in rapid extravasation of the contrast material appeared as early Time2Max and typically circumferences the tumor periphery (Fig. 3I)
. All of the VEGF overexpressing C6-pTET-VEGF tumors (n = 14) showed typical radial convection of contrast material around the tumor, with regions further from the tumor boundary that showed low initial rate of accumulation of contrast and gained high accumulation rate at later time points (Fig. 2)
.
We additionally checked the correlation among tumor diameter, interstitial flow velocity, and permeability. Interstitial flow velocity was calculated as the rate of change in the average radius of the enhancement ring between 10 and 60 min after contrast. APS and interstitial flow velocity of biotin6-BSA-GdDTPA26 in C6-pTET-VEGF tumors correlated significantly with tumor diameter (Fig. 5
; t test, P = 0.004; r2 = 0.52 for interstitial flow velocity, and P = 0.05; r2 = 0.28 for APS; n = 14). Switching off VEGF overexpression in C6-pTET-VEGF tumors resulted in substantial inhibition of vascular permeability (Fig. 3K
; Fig. 4B
) and accordingly interstitial convection was not evident in Time2Max maps (Fig. 3L)
.
|
-SMA, whereas mature vessels were abundant in the tumor rim but could be observed also within the tumor (Fig. 6, A, B, E, and F)
. The biotin tag was used for histological staining of the MR contrast material (Fig. 6, C, D, G, and H)
. VEGF overexpression was associated with significant extravascular accumulation of the contrast material in the peritumor interstitial region, whereas in tumors subjected to VEGF withdrawal, the contrast material was confined to blood vessels in the peritumoral area (Fig. 6, C, D, G, and H)
. Thus, histological staining for the biotinylated contrast material confirmed the MRI results.
|
. Lymphatic labeling was resolved in about half of the examined C6-pTET-VEGF tumors (n = 9 of 16) with up to three draining lymphatic vessels for each tumor. Lymphatic uptake was also observed in 1 (of 3) C6-pTET-VEGF tumor after VEGF withdrawal as well as in 2 (of 4) C6-pTET-VEGF tumors treated chronically with tetracycline.
|
. To separate between permeability and lymphatic drain, and to map drain at early time points, we used the biotin tag on the contrast material and applied an in vivo "avidin chase" (n = 6; Fig. 8, D–I
). Biotin6-BSA-GdDTPA26 was injected i.v. and was allowed to extravasate for 30 min (Fig. 8D)
. i.v. avidin administration eliminated almost all of the intravascular biotinylated contrast material within 3 min (Fig. 8E)
. One h after avidin chase the contrast was no longer evident in the blood (Fig. 8F)
. During that time extravasated biotin6-BSA-GdDTPA26 showed interstitial convection and dilution. Signal enhancement was also apparent at the vicinity of the popliteal lymph node (Fig. 8, E–I)
. Second administration of biotin6-BSA-GdDTPA26 was done to rehighlight the blood vessels at the end of the MRI session (Fig. 8G)
.
|
and 10
). Between these scans the mice were allowed to wake up and walk freely so as to facilitate the natural lymphatic function. At 6 h these images showed significant clearance of the contrast material from the blood vessels, whereas the contrast material that was observed extravasating at the tumor during the first hour drained from the skin into inner tissues (Fig. 9)
. Vessel-like structures, delineating the drainage pathway, were detected (Fig. 10)
. Readministration of a second bolus of biotin6-BSA-GdDTPA26 was done to highlight the blood vessels relative to the distribution of the extravasated contrast at 6 h (Fig. 9, D and E)
. Selective slices from the three-dimensional datasets showing the blood vessels (in red) and the extravasated contrast (in green) were overlaid on a grayscale MIP. These composite images demonstrate that the clearance of the labeled albumin reveals vascular structures that are distinctly not blood vessels, and lead to the general direction of the popliteal lymph node (Fig. 10)
. A similar drainage pattern was observed also at later time points (6–12 h; n = 4).
|
|
|
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Avidin chase was applied here for generating experimental separation between vascular permeability and subsequent drain. Avidin, as well as avidin complexed to the chased biotin-protein, are cleared from the blood stream to the liver within a few minutes, and thus, avidin can be used for inducing rapid clearance of biotin-tagged molecules (25) . Avidin chase has been applied previously in nuclear medicine to improve specificity of radiolabeled antibodies (20 , 26) . To the best of our knowledge this is the first demonstration for MRI application of avidin chase. By application of avidin chase we induced rapid clearance of intravascular biotin6-BSA-GdDTPA26 as detected by MRI. This allowed detection of the directional convection and draining of the contrast material extravasated from the tumor, which is induced by the continuous leak of fluids and plasma proteins (including unlabeled albumin). The draining from the tumor can be followed by MRI to depths and distances that exceed by far those regions accessible by intravital optical methods (27) .
Switching off VEGF expression decreased fPV and abolished permeability. Under these conditions, interstitial convection could not be detected. The decrease in fPV may reflect vasoconstriction and destruction of tumor vasculature in response to VEGF withdrawal as we demonstrated previously by histology and by blood oxygenation level dependent (BOLD) contrast MRI (18 , 19 , 28) . The complete shutdown of permeability on acute VEGF withdrawal suggests that the environmental conditions were not sufficient to activate the endogenous, hypoxia-regulated expression of VEGF (29) . These results are consistent with reduced permeability reported previously in MRI studies of tumors treated with anti-VEGF antibodies (21 , 22) .
VEGF overexpression increased the probability of lymphatic uptake in the skin layer. However, lymphatic uptake was also observed in some cases after blocking VEGF overexpression with tetracycline. The few skin lymphatic vessels detected by confocal microscopy might be because of a number of reasons. Close to the tumor the interstitial concentration of the contrast material was high, and the concentration in lymphatic vessels will equilibrate with that of the interstitium. Thus, the vessels will not be detectable by fluorescence microscopy. Further from the tumor the drain will be more efficient into deeper and larger collecting vessels than in the superficial skin layer. Lymphatic vessels draining the skin into deeper tissue are damaged during tissue excision and are beyond the limit of light penetration. Therefore, information coming from connecting and deeper vessels might be lost.
Optimal lymphatic imaging should provide three-dimensional information along with high spatial resolution. Three-dimensional GE MRI scans were performed 6–12 h after contrast administration. By that time, the concentration of the labeled albumin in the blood was significantly reduced, and the extravasated contrast material was drained from the tumor showing tracks of vessel-like structures. Alternatively, we demonstrated that intravascular contrast material could be chased by avidin allowing separation of intra- and extravascular contrast material, and enabling direct and continuous MRI tracking of the convection and lymphatic drain of the contrast material that extravasated in the tumor at earlier time points.
Two recent studies reported that tumor-induced lymphangiogenesis in VEGF-C overexpressing tumors could be suppressed by soluble extracellular domain of the VEGF-C receptor, VEGFR-3, or soluble fusion protein, VEGFR-3-immunoglobulin (4 , 30) . In both cases VEGF-C was associated with lymphangiogenesis, but invasion of lymphatics seems to require additional regulating factors and mechanisms (4 , 30) . We demonstrated here that VEGF induced not only vascular permeability and increased vessel density, but also interstitial convection and lymphatic drain. Thus, this study demonstrates a direct link between tumor expression of VEGF and peritumor lymphatic activity, suggesting that VEGF-induced permeability could impact lymphatic-mediated metastasis. The entire panel of responses to VEGF, including increased vascular permeability, blood-plasma volume, interstitial convection, and lymphatic drain, could be followed in vivo by MRI after a single or double i.v. administration of tagged albumin. Furthermore, vascular permeability could be resolved experimentally from lymphatic drain using in vivo avidin chase.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
1 Supported by Grants NIH RO1 CA75334 (to M. N.) and RO1 CA90471 (to M. N. and Z. M. B.). Its contents are solely the responsibility of the authors and do not necessarily represent the official views of Weizmann Institute of Science, the Johns Hopkins University, or NIH. 
2 To whom requests for reprints should be addressed, at Department of Biological Regulation, The Weizmann Institute of Science, Rehovot, 76100 Israel. Phone: 972-8-9342487; Fax: 972-8-9342487; E-mail: michal.neeman{at}weizmann.ac.il
3 The abbreviations used are: VEGF, vascular endothelial growth factor; GdDTPA, gadolinium-diethylene triamine pentaacetic acid; VEGFR, vascular endothelial growth factor receptor; MRI, magnetic resonance imaging; MR, magnetic resonance; TR, repetition time; TE, echo time; FOV, field of view; fPV, blood-plasma volume fraction; APS, apparent permeability surface area product; SMA, smooth muscle actin; MIP, maximal intensity projection; GE, gradient echo (image).
Received 7/ 1/02. Accepted 9/20/02.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  A. A. Tomei, S. Siegert, M. R. Britschgi, S. A. Luther, and M. A. Swartz Fluid Flow Regulates Stromal Cell Organization and CCL21 Expression in a Tissue-Engineered Lymph Node Microenvironment J. Immunol., October 1, 2009; 183(7): 4273 - 4283. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S.-F. Chang, C. A. Chang, D.-Y. Lee, P.-L. Lee, Y.-M. Yeh, C.-R. Yeh, C.-K. Cheng, S. Chien, and J.-J. Chiu Tumor cell cycle arrest induced by shear stress: Roles of integrins and Smad PNAS, March 11, 2008; 105(10): 3927 - 3932. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. J. M. Mulder, D. W. J. van der Schaft, P. A. I. Hautvast, G. J. Strijkers, G. A. Koning, G. Storm, K. H. Mayo, A. W. Griffioen, and K. Nicolay Early in vivo assessment of angiostatic therapy efficacy by molecular MRI FASEB J, February 1, 2007; 21(2): 378 - 383. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. P. Pathak, D. Artemov, M. Neeman, and Z. M. Bhujwalla Lymph node metastasis in breast cancer xenografts is associated with increased regions of extravascular drain, lymphatic vessel area, and invasive phenotype. Cancer Res., May 15, 2006; 66(10): 5151 - 5158. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Mazooz, T. Mehlman, T.-S. Lai, C. S. Greenberg, M. W. Dewhirst, and M. Neeman Development of Magnetic Resonance Imaging Contrast Material for In vivo Mapping of Tissue Transglutaminase Activity Cancer Res., February 15, 2005; 65(4): 1369 - 1375. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. P. Pathak, D. Artemov, B. D. Ward, D. G. Jackson, M. Neeman, and Z. M. Bhujwalla Characterizing Extravascular Fluid Transport of Macromolecules in the Tumor Interstitium by Magnetic Resonance Imaging Cancer Res., February 15, 2005; 65(4): 1425 - 1432. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N Raghunand, R A Gatenby, and R J Gillies Microenvironmental and cellular consequences of altered blood flow in tumours Br. J. Radiol., December 1, 2003; 76(suppl_1): S11 - S22. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Israely, H. Dafni, D. Granot, N. Nevo, A. Tsafriri, and M. Neeman Vascular Remodeling and Angiogenesis in Ectopic Ovarian Transplants: A Crucial Role of Pericytes and Vascular Smooth Muscle Cells in Maintenance of Ovarian Grafts Biol Reprod, June 1, 2003; 68(6): 2055 - 2064. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cancer Prevention Research |
| Cancer Prevention Journals Portal | Cancer Reviews Online |
| Annual Meeting Education Book | Meeting Abstracts Online |
; n = 4; 3 vessels were studied in two to three image slices in each mouse) and by biochemical determination of blood biotin in blood samples taken from the tail (
; n = 6) and the heart (
; n = 3). Data are presented as percent change relative to initial concentration (mean value); bars, ±SD.
), and late enhancing draining tissue (10–60 min after contrast; 
